Sunday, September 18, 2016

Markets: Volatility Ahead

Finally we had a little bit of excitement back in the markets, and a vindication of sort for the numerous bears. Analysts from Citi confirm the macro drivers to blame. Indeed the cross market correlation has been on the rise. Below table shows the current cross market correlation1. As it shows, rates and inflation still remain the major drivers, along with a very high level of correlation between commodities and currencies.

The chart below shows average correlation across asset classes. The recent spike is still far away from the levels around August 2015, but clearly captures the market sentiment.

And this sentiment is what is reflected in the latest AAII release last week. The market neutrals and bears remain significantly above the historical average as we have during most of this post-crisis bull runs. The change to highlight is an increase in bears at the expense of mostly the neutrals. 

For a contrarian, this would signal a limited scope of continued sell-off. However the other factor is positioning on the derivatives side. I have written about this before, but you must have already observed the change in the intraday price patterns in S&P. During much of last few months, it showed a strong mean-reverting characteristics (opening price shocks reversed during the day). For last few session starting from the 9th sell off, it seems the intraday trends are self sustaining now. That is confirmed on more quantitative measures as well. The chart 2 below shows two approximate indications of short gamma positioning of the dealers. The idea behind this is in a market where the dealers (i.e. the hedgers, as opposed to players who hold options positions unhedged) are short gamma (net sellers of options), the natural hedging activity will create price pressure that will tend to amplify a price move (a up move in to a sustained rally and vice versa). On the other hand, when the dealers are net long, this will tend to stabilize price moves. Much of the stability in S&P intraday move for past few months can, at least partially, be attributed to net long position from the dealers, which appears changing now, as the short term intraday trends get stronger and sustain longer.

On the valuation side, however, S&P is still not screaming over-valued. Below chart shows world equity markets valuation vs trends (past 1-year returns) in two measures. The left one is the regular P/E measure, on which S&P is quite in the red zone, along with India and only second to Mexico. However, just going by historical P/E in a world of zero rates can be highly misleading. In terms relative valuation to bonds, S&P is quite in the middle.

If you followed the trades from my last post, it would have been a good couple of weeks capturing most of the major moves in the markets in the right direction. Looking ahead, if you are a bear, the investor sentiments and the valuation is not at a very helpful support to go big short at current levels. On the other hand the change in the gamma signature of the markets tells us unless something changed after Friday's expiry, we will continue to see decent swings and volatility will pick up. Although vols are not particularly cheap (relative to realized, yet), and I think given the reasons discussed before, it still makes more sense to buy options than to buy VIX here.

A traders' market after a long time. Brace for the upcoming Fed, but more for the BoJ. Going beyond equities, the major moves in rates (one of the major driers across asset classes now) has been set in motion by BoJ arguably. The recent bout of steepening in fixed income started with a bout of sell-off in JPY rates markets. This transmitted to rest of the world following ECB, with sharp steepening across EUR, USD, and GBP. The built up to the month end BoJ is almost palpable, and if not FOMC, at least this is almost certain to be an interesting event.

1. This is based on smoothed data (Gaussian kernel smoothed with 5-day bandwidth) to capture medium-term correlation.
2. The left chart is based on identifying trends quantitatively using change point techniques. The idea is as trends become more sustaining the ratio of max to median trends will increase, as shown in the chart. The right hand chart shows absolute value of beta in a simple regression of intraday price to time, capturing the strength of the trend (if any) irrespective of its direction (rally or sell-off).
3. Data from Bloomberg and Google Finance